Probabilistic Kernel Function for Fast Angle Testing

Kejing Lu, Chuan Xiao, Yoshiharu Ishikawa · 2025

Read Probabilistic Kernel Function for Fast Angle Testing by Kejing Lu, Chuan Xiao, Yoshiharu Ishikawa (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2505-20274.txt.

Description

Kejing Lu, Chuan Xiao, and Yoshiharu Ishikawa speed up angle testing for similarity search in 2025. Finding near neighbors in high-dimensional data often hinges on angles between vectors. Random projection is the standard trick, but it leans on many random vectors.

Their kernel functions use fixed reference angles and a set projection structure instead. No large-sample assumptions are needed for the method to hold. Tests in theory and practice beat the classic random method. Used in a neighbor search task, their approach served two to three times more queries per second than a leading graph method.

Catalog record

CREATORKejing Lu, Chuan Xiao, Yoshiharu Ishikawa
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE63 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2505-20274

Browse the full library · About the Tower of Babel